Cloud & AI Efficiency Management in Practice
Cloud & AI Efficiency Management represents a practical discipline for systematically ensuring the efficient utilization of cloud-native and third-party services. Rather than treating cost optimization as a one-time exercise, it establishes ongoing processes that detect waste, provide context, and enable rapid remediation.
This session, featuring Taylor Houck from PointFive and Mike Graff from Dolby, examines what this looks like in practice -- particularly when applied to Snowflake, one of the fastest-growing cost centers in modern data stacks.
Real-World Snowflake Optimization
The discussion draws on Snowflake optimization strategies taken directly from enterprise deployments. Snowflake's consumption-based pricing model creates unique optimization challenges: warehouse sizing, auto-suspend configuration, query efficiency, and storage management all interact in ways that can dramatically affect costs.
Common Snowflake inefficiency patterns include:
- Over-provisioned warehouses running at low utilization during off-peak hours
- Queries that scan excessive data due to missing clustering keys or poor partition pruning
- Auto-suspend settings that are either too aggressive (causing cold-start latency) or too lenient (wasting credits on idle compute)
- Storage costs from unmanaged time travel retention and transient table policies
Understanding these patterns and detecting them automatically is where Cloud & AI Efficiency tools provide the most value -- surfacing specific, actionable recommendations rather than generic advice.
The Efficiency Hub: A Community Resource
The session also previews the Efficiency Hub, an open, community-driven resource designed to catalog inefficiencies across major cloud platforms. The Hub serves as a shared knowledge base where practitioners can contribute and access optimization patterns, helping the broader community move beyond ad-hoc fixes toward systematic efficiency practices.
By making inefficiency patterns publicly accessible and community-maintained, the Efficiency Hub aims to accelerate the maturity of cloud optimization practices across the industry.